The declining role of ethnicity in farm householddifferentiation: A case study from Ngoc Phai

Commune, Cho Don District, Bac Kan Province,Viet Nam

Jean-Christophe Castella a, b, Tran Ouoe Hoa b,

Olivier Husson b, C, Vu Hai Nam d, Dang Dinh Ouang b

a Institut de Recherche pour le Développement (IRD),

213 rue Lafayette, 75480 Paris Cedex 10, France, and

International Rice Research Institute (IRRI), DAPO 7777, Metro ManHa, Philippines

b Mountain Agrarian Systems (SAM) Program, Vietnam Agricultural Science Institute (VASI),

Thanh Tri, Ha NOi, Viet Nam

C Centre de Coopération Internationale en Recherche Agronomique pour le Développement

(CIRAD), Av. Agropolis, F34398 Montpellier Cedex 5, France

d GIS and Remote Sensing Center, National Institute for Agricultural Planning and Projection

(NIAPP), 61 Hang Chuoi, Ha Noi, Viet Nam

Abstract

Farming has changed dramatically over the last fi fty years in Ngoc Phal Commune, as inail of Bac Kan Province, Historically, ethnicity has long been the key determinant ofaccess to land in northern Viet Nam uplands, and thus the determinant of farmers'agricultural practices. However, major external policy changes have periodically alteredthe environment in which farmers plan their livelihood strategies. Agriculturalcooperatives were established throughout the country in the 1960.'s, and then dismantledsorne twenty years later, returning the land to individual family farms. During thecooperative period, ail ethnie groups contributed to the intensification of paddylandproductivity, thereby limiting agricultural pressure on the uplands. Nonetheless, riceproduction was not sufficient to coyer food needs because of management problems in thecooperatives, Therefore, upland rice production became indispensable to meet the deficitin paddyland production. In the 1990s, the allocation first of paddyfields and later ofuplands helped to slow the deterioration of the mountain ecosystem. However, the localreinterpretations of national land polities resulted in further inequalities, placing certainethnie groups into extreme poverty and food insecurity. Tày farmers who couId reclaimancestral paddy lands were often privileged over the Dao, who traditionally had beenshifting cultivators and thus were forced to return to slash-and-burn cultivation systemsthat were no longer sustainable under higher population pressure and a new institutionalenvironment.

J.C. Castella and Dang Dinh Quang eds. (2002) Doi Moi in the Mountains. Land use changes and farmers'/ive­/ihood strategies in Bac Kan Province, Vietnam. The Agricultural Pub/ishing House, Ha Noi, Viet Nam. 47-71.

J. C. CastelJa, Tran Quoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Quang

Ethnicity was a major detenninant of livelihood strategy in the past, and contributedgreatly to the household differentiation that exists at present. However, distinctions amonglivelihood strategies can no longer be drawn along ethnie lines. Fanners of ail ethnicitiesshare the goal of attaining food security through paddyland rice production. Fanners wholack paddyland fields, whether Tày or Dao, are tuming to shifting cultivation on thehillsides.

Keywords: mountain agriculture, farming-systems typology, land use changes, crop-livestockinteractions, household differentiation, northem Viet Nam

1. Introduction

The unsustainable use of forest resources in combination with rising populationpressure endangers the ecological and economic sustainability of traditionalupland agrarian systems (Bal et al., 1997). High population pressure has madeaccess to land that is suitable for agriculture a vital issue for subsistence fanninghouseholds (Castella and Erout, 2002).

Vietnam Govemment authorities have long blamed northem Viel Nam's reductionin forest coyer on ethnie minority groups, such as the Dao and the H'mong, thatpractice slash-and-bum shifting cultivation (Dang Ng1::ljem Van, 1991; MOITisonand Dubois, 1998). The Dao within Ngoc Phai Commune, despite having adoptedfixed settlements since 1962, are included in the above group blfause theycontinue to use slash-and-bum practices. '

The Tày ethnies were the first settlers in Ngoc Phai Commune, and now controlthe majority of the paddyfields there. However, it would be an oversimplificationto view Dao agriculture as confined to the uplands and Tày agriculture to thelowlands. There are Tày who have been without paddyfields since the landreallocations of the early 1990s, and there are Dao who have gained access tolowland paddyfields through the recently-created land market in Ngoc Phai.

To assist households in developing production systems that are moreenvironmentally sustainable, we must first understand the processes andcircumstances that have led to people's CUITent practices. We also mustcharacterize the diversity of existing household production systems to ensure thatour proposed technical and organizational innovations are adapted to the localcommunity's objectives and constraints. Towards this end, this chapter presentstwo components ofa monographie study ofNgoc Phai Commune: (i) an historicalanalysis of commune-level agrarian dynamics and their environmental impact,and (ii) a fann-household differentiation study.

48

RaIe af ethnicity in farm hausehald differentiatian

2. Methods

2. 1. Selection of Ngoc Phai Commune

Cho Don District, situated approximately 220 km north of Ha Noi, is one of sixdistricts of Bac Kan Province. The total area of the district is approximately92,000 ha and the population 46,800, for an average population density of 51people 1 km~. The Tày ethnie group makes up 76% of the population. The otherethnie groups, in order of size, are the Dao (10%), the Kinh (10%), the Nung (2%),and the H'mong.

Within Cho Don District, we selected Ngoc Phai Commune for our monographiestudy based on the agro-ecological zoning of Cho Don District (Castella et al.,2002b). The commune captures the diversity of landscapes, ethnie groups, andagronomie issues present in the district (Bal et al., 2000).

2.2. Relating farming systems to land-use changes

The overall approach was based on relating farming systems to land-use changes.There were four main steps in the study:

- The agro-ecological zoning of the district, on the basis of which the Ngoc PhaiCommune was chosen for the study. This was based on the analysis of secondarydata (maps, statistics, etc.);

- An examination of agrarian system dynamics over the last five decades, withinformation gained from surveying witnesses to the commune's history;

- An examination of land-use changes, with the use of a chronological series ofaerial photographs (1983, 1989) and SPOT satellite images (1990, 1995 and 1998);

- Surveys ofa sample offarmers (n = 250) representative of the diversity of groupspresent in the commune. Using a semi-structured questionnaire, we quantifiedthe land use of individual farms and their associated options for futureproduction systems. This allowed us both to explain the present diversity and toidentify patterns of future agriculturalland use evolution.

3. A tiered system of land use and access

3.1. Tày - Dao: a dual agrarian system (up to 1960)

The Tày and the Dao, historically the two dominant ethnie groups in the studiedregion, have traditionally occupied different tiers of the ecosystem. These twoethnie groups have distinct cultural norms and practices, and the evolution patternsof their land use systems have followed unique courses, even if shaped by thesame State policies.

49

Table 1: Comparison of the traditionsl T~y and Daocultivation systems

Ethnie Group Tày Dao

System Type Sedentary Shifting

Location Lowlands Uplands

Main crop Irrigated rice Upland rice

Limiting factor ofLabor Weeds

upland field area

Type ofupland fallowShort (5-10 Long (20

years) years)

Importance of forestMedium High

products in diet

Role ofbuffaloesPlowing, Capital,fertilizer transport

J.C. Castella, Tran Quoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Quang

Until the establishment of cooperatives in 1960, the Tày inhabited the lowlandsand the Dao inhabited the uplands. The Tày cultivated an annual irrigated ricecrop, using a collective system of water and labor management based on mutualaid. The plentifu1 lowland area met the needs of a still-small population (density< 1°people / km2

). On the gentie slopes bordering the lowlands, Tày farmerscultivated maize and cassava as feed for pigs, or less commonly as a securityagainst food shortages from poor rice harvests or natural disasters. Sorne familiesbuilt terraces for rice production on sloping land, despite the availability oflowlands and the burden ofsuch construction. Buffaloes were raised both for draftpower and manure for fertilizing paddyfields. Social differentiation was based onland ownership; early arrivais to a village tended to gain possession of the largestand most productive paddyfields.

The Dao, the "people of the forest," were primarily producers of upland rice, butalso cultivated maize and cassava. Their shifting cultivation system was based onswidden agriculture with long fallow periods, necessitating frequent migrations.The gathering of forest products (bamboo shoots, wild vegetables, mushrooms,etc.) and hunting of still-plentiful game played essential roles in the Dao system,particularly in times of food shortage. Fields were grouped together to limit croplosses to wild animais, but the typical hamlet consisted of only two to fourhouseholds. Upland territory was subject neither to Tày-like land regulations norto taxes levied by the French colonial administration. A plot belonged to thefamily that had cleared it for the duration of use, but became available for otherusers as soon as it was abandoned. Given the lack of permanent property, landownership was not a criterion for social differentiation. The socially elevated in aDao community were those whohad water buffaloes (up to sixper family), which served as akind of savings as weil as beinguseful for transportation whenhamlets were being relocated.

Interactions between the Daoand the Tày were rare, as theirrespective production systemswere based on very distinctlandscape units. Table 1 pre­sents the main characteristics ofthe pre-cooperative systems.With Viet Nam's independencein 1954, the traditional systembegan to see growing inter­vention by the State.

50

Role of ethnicity in farm household differentiation

3.2. fnstitutionalizing mutuaf aid

The first land reforrn occurred in 1954, and aimed to create equality in rural areasby confiscating land from wealthy landowners and giving it to the mostimpoverished. In Ngoc Phai, as in ail of Cho Don District, the reforrn had only aminimal effect because of the low level of social differentiation (Tran Van Ha,1993). In the lowlands, agricultural officers oversaw the creation of mutual aidgroups composed of an average of five households each. At this stage, only theTày participated. Thus, the mutual aid groups effectively institutionalized thetraditional Tày system of mutual aid for water and labor management. The officialmutual aid system made possible an increase in irrigable land area, which resultedin increased productivity.

3.3. The agricufturaf cooperatives

In 1960, the first cooperative, ten households in size, was in Phieng Lieng village,populated by the Tày. Its success encouraged ail the other villages in Ngoc PhaiCommune to organize themselves as cooperatives. Membership was voluntary,but with time the free-market system dwindled away, and it would have beendifficult for any individual to exist separately from a village cooperative.Nonetheless, there were villages in Bac Kan Province that did not becomecoopemtives.

In 1961, a national program attempted to settle the country's ethnie minoritiesperrnanently. The goals of the program were to settle the ethnie minorities inregions where they could contribute to cooperative production, and also be moreeffectively observed and controlled by local authorities. This program placedeighteen Dao families into Tày-populated Ban Cuon village, on the terraced lowerslopes. Sorne Tày families took the opportunity to leave Ban Cuon and join otherTày-dominated cooperatives, while others stayed to work alongside the Dao in thevillage. In spite of the departures of sorne Tày families from Ban Cuon, the arrivaiof Kinh migrants from the overpopulated Red River delta resulted in substantiaJnet population growth in the area. Between 1964 and 1974, more than 4,000 Kinhpassed through Cho Don District to build the road connecting Bac Kan and Ba Betowns, and sorne stayed behind after the work was finished. The result was amigration of the immigrant K[nh toward the interior of the commune andincreased pressure on commune paddyfields (Figure 2).

In 1962, Ngoc Phai Commune was composed of two distinct cooperatives: thepredominantly Dao cooperative in Ban Cuon village, and the Ba Ngoc cooperativecontaining the other five Tày villages. Though founded on prin-ciples ofegalitarianism, the cooperative system resulted in a marked diffe-rentiation ofvillages based on per capita ricefield area. Despite low population density (Figure1), the retum per person per day on a paddyfield was fairly unequal, ranging from0.8 kg in Ban Cuon Cooperative to 1.5 kg in Ba Ngoc Cooperative. Like the

51

J.C. Castella, Tran Quoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Quang

Figure 1: Changes in population density in Ngoc PhaiCommune

2000

2000

1990

1990

+ Commune

• Tày Village

• Dao Village

1980

1980

1970

1970

+ Commune

• TàyViliage

• Dao Village

1960

90807060504030­2010O-'--~--~--~----~---l

m2, person

3500 ,-----------------... --- ------.....---....--.-,

3000

2000

2500

Persons , km2

Source: interviews and secondary data

1500

1~~~1 -: ~ ;1960

Source: interviews and secondary data

traditional livelihood systems,the new system was split alongethnie lines: the Dao coopera­tive fanned an average paddy­field area of 550 m2 per person,compared to 2600 m2 in the Tàycooperative (Figure 2).

As the Tày cooperativeintensified its rice cultivationsystems (by introduction of tworice crops per year, expansion ofirrigated area, new photoperiod­insensitive short-cycle varieties,use of chemical inputs, plowinnovations, etc.), paddy yieldswere stagnating and evendeclining in the Dao cooperative(1.7 t/ha compared with morethan 5 t/ha in the Tàycooperative). Lowland paddy­field production did not meet thefood requirements of Ban Cuonvillage, leading the Daocooperative there to turn to theuplands. The remunerationsystems differed substantiallybetween lowland and uplandcultivation: a large proportion(30%) of lowland productionwent to the State, whereas thefull production of the uplandswas shared among cooperativemembers. Further, swidden

fields cleared from primary forests offered higher rice yields: 4 t/ha in swiddenfields compared to lowland yields of 1.7 t/ha/year (1 crop cycle; Dao) and 5.0t/ha/year (2 crop cycles; Tày), and higher labor productivity: 32,000 VND/day inswidden versus 20,000 VND/day in lowland paddyfields, both in 1999 VN Dongs.With work on swiddens considerably more profitable than on paddyfields, it is notsurprising that the Dao did not invest in intensifYing paddyfield productivity.

As disparities between the cooperatives progressively deepened, new disparitiesarose among individual households within the Dao cooperative. Sorne householdsin the Dao cooperative developed private swidden fields alongside the collective

Figure 2:Changes in paddyfield area per persan inNgoc Phai Commune.

52

RoJe of ethnicity in farm househoJd differentiation

swidden fields. The blurred borders between collective and private sloping landsmade it easy for households with large labor forces to engage in more-profitableprivate activities (upland rice, pig raising, forest product extraction, etc.) once theyhad completed their required collective work.

In contrast, the Tày cooperative did not develop private production. Theyachieved food security through ricefield intensification, and their demandingricefield work schedules did not leave enough time for the cultivation of swiddencrops. In summary, a double process of differentiation occurred: (i) betweencollectives, based on per capita paddyfield area; and (ii) among households withinthe Dao cooperative, based on the ratio between a household's labor force andnumber ofmouths to feed (Sadoulet et al., 2002).

3.4. Crisis in the 1970s

In 1970, the two village cooperatives were merged into one large commune-levelcooperative. The new level of management brought bureaucratization andreduced flexibility, leading to a growing lassitude on the part of cooperativemembers. Chemical input allocations were poorly organized, and yields droppedas low as 2t/ha/cycle in spite ofthe cooperatives' successes with Green Revolutiontechnologies before they merged.

Exacerbating the situation, the commune population had doubled in the precedingtwenty years (Figure 1). Population growth exceeded the rate of new paddyfieldconstruction, causing paddyfield area per person to drop by approximately 40%(Figure 2). The retum on labor used in paddy production was steadi1y declining,and cooperative work could no longer meet househo1ds' needs. Therefore,farmers of all ethnicities tumed to the hillsides, in spite of bans imposed (but notyet enforced) by authorities. Specifically, a 1975 forest-protection law restrictedthe right to harvest wood to State forestry enterprises, and launched the first forestplantation programs, but was received with general disinterest by the Ngoc Phaipopulation. Slash-and-burn practices were made punishable by severe fines,resulting in a food crisis between 1977 and 1980, bringing Ngoc Phai to the edgeof famine.

3.5. Decree 100, the first step toward decollectivization

In response to similar food crises throughout the country, the State passed Decree100 in 1981, allocating paddy fields for 4-5 years to househo1ds based on thenumber offami1y members. Households owed a fixed amount of their productionto the cooperative, based on the size and quality of their fields, but were free toretain any surplus for themselves. Families that had not owned cattle before thecollective period cou1d now sell sorne rice and use the income to take advantageof low priees (50% of animais' cooperative-period value).

53

Table 2: Land use strategies (types A and B)developed by households after Deeree 100

N.B.: The (+) sign represents an increased value andthe (-) sign a deereased value ofan indicator.

Household type A B

Ratio of laborers to number of mouths+

to feed per household-

Ricefield area per laborer - +

Swidden area per laborer + -

Production strategy- Cash crops + -- Maize and livestock + -- Land accumulation + -

Capital accumulation + -

J.C. Castella, Tran Quoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Quang

In Ngoc Phai, the new regulations had effects opposite from what was intended.Paddyfield productivity did not increase substantially because of continuedproblems with cooperative management, for example with allocations of chemicalinputs. Labor continued to be more profitable on swiddens than on paddy fields,and hillside cultivation dominated, to the detriment of paddyfield intensification.In 1983, the high profitability of swiddens led to an uncontrolled rush by Tàyfarmers to claim as much forestland as possible. Within five years, swiddencultivation had claimed ail of the commune's clearable land. Plots were often sofar from the villages that farmers had to live in temporary bamboo shelters duringthe cultivation season.

The Tày and Dao villages again tumed to divergent land uses. In the Dao villages,sloping lands were used primarily for upland rice. Production exceededconsumption needs, and farmers stored upland rice surpluses in granaries forfuture security, but sold surplus irrigated rice. Tày farmers likewise used sorneuplands to grow sticky rice, but treated their upland rice as a cash crop. The Tàydid not stockpile their rice surpluses; both upland and lowland surpluses weresold. In addition, Tày farmers used sorne swidden fields to grow maize for pigraising, which was rapidly developing thanks to the introduction of faster-growinghybrid breeds. Hybrid pigs required considerably richer diets than their traditionalcousins, hence the increase in land area devoted to maize.

The period following Decree 100 led to significant household differentiationwithin villages in Ngoc Phai Commune (Sadoulet et al., 2002). Paddyfields hadbeen distributed based on the total number ofindividuals in a household, whetherproductive (laborers) or non-productive (children, elderly, or invalids).Households with greater pro­portions of laborers could morequickly complete their man­datory lowland work and engagein more profitable activitiessuch as swidden cultivation. lncontrast, households with smallproportions of laborers had tostruggle to meet their workobligations, and faced penaltieswhen quotas were not achieved.Table 2 summarizes the diffe­rences in relative land and laborendowment, and resuitingstrategies, between the twohousehold types that developedduring this period.

54

Role of ethnicity in farm household differentiation

During the Decree 100 period, increasing private production (mostly extensiveslash-and-burn agriculture) resulted in (i) better nutrition for farmers, (ii) thebeginning of capital accumulation, (iii) a notable deterioration offorest resources,and (iv) differentiation among households.

3.6. Resolution 10, the second step toward decol/ectivization

The disparities caused by Decree 100 prompted the State to pass Resolution 10 in1988, bringing about the dismantling of the cooperatives. Farmers gained controlof aIl cropping activities, chemical input allocation, and irrigation. They regainedaIl of the means of production except the land, for which they were granted usagerights by the State. Specifically, paddyfields were redistributed in proportion tothe number of laborers in each household. Each household was free to use aIl theproduction from their land however they chose. The family farm became theelemental unit of production and a free, private market developed includingpurchase/sale of rice, fertilizer, equipment such as huskers, and buffalo meattargeted at Ha Noi consumers.

The latest in a long series of policy changes, Resolution 10 finally was able tostimulate increased paddyfield production, helped by the declining profitability ofslash-and-burn production. Farmers began to invest more time and capital inpaddyfields, and yields per crop increased. In addition, a substantial proportion ofpaddy fields passed from one crop per year to two.

At this time, a group of Tày founding families (i.e. historically the first settlers inthe area) in neighboring Cao Bang Province began a movement to reclaimancestral lands, a movement that quickly spread through aIl ofBac Thail

• Farmerswho had not been members of the pre-1954 Tày mutual-aid groups were divestedof almost ail of the lowland plots that they had been farming since the creation ofthe first cooperatives in 1960. The ownership of ricefields that had been builtduring the cooperative period was settled in negotiations. Land conf1ictsoccasionally arose but the reallocation process was irreversible, and households inNgoc Phai soon differentiated based on access to paddyfields. After the foundingfamilies reclaimed their ancestral lands a dual agrarian system emerged againbased on the following two household types:

The Tày founding families. By 1991, the founding Tày families had claimedalmost aIl paddyfields in Ngoc Phai. Most were growing two crops per year, andthe increasing availability of chemical inputs along with the use of organicfertilizers (pig and buffalo manure) contributed to a further increase in yields (upto 2.8 t/ha/cycle). For these well-endowed households, the period was marked bythe intensification of irrigated cultivation in the lowlands and the progressive

1Bac Kan Province was created in 1997 from the joining oftwo districts in Cao Bang

Province and a part of Bac Thai Province.

55

J.C. Castella, Tran Ouoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Ouang

abandoning of upland rice. Labor productivity on sloping land (8,000 VND/day)had become considerably lower than on the paddyfields (20,000 VND/day).

The landless. Regardless of their ethnie heritage, families who were livingoutside of their native villages were deprived of their paddyfields. In Ban Cuon,the Dao community retained only one hectare of their paddyfields, a field that hadbeen built during the cooperative period. The combination of increaseddemographic pressure and the loss of paddyfields forced the Dao to expand theirswidden areas, and the search for fertile forestland led them weil beyond theborders of the village. But with population density now hear 29 people/km2

,

forests old enough to provide swidden yields comparable to pre-cooperative times(2t/ha) were difficult to come by (Husson et al., 2001). Farrners increased thenumber of successive years of cultivation and decreased the length of fallowperiods (Castella et al., 2002).

Sorne families chose at this time to migrate to southem Viel Nam, drawn by theintensive development of coffee crops on pioneer lands (De Koning, 1999; Altheret al., 2002). Others, mostly of Kinh origin, opened small shops along the roads.Poor households tumed to the extraction of forest resources, which provided anecessary complement to their income.

3.7. The 1993 land law

The 1993 land law ratified the land redistribution process that had already begun.Use rights both for flatlands with annual crops (paddyfields) and for aquacultureponds (fish farrns) were allocated for periods of 20 years, whereas use rights forforestland were allocated for periods of 50 years, though all land was stillofficially owned by the State. Farrners could now transfer, sell, buy, rent, andinherit land-use rights.

After almost 10 years of changing land-access mies, clearly defined use rights andaccess to paddyfields and forests caused new production strategies to develop (cf.Section 4). The new mIes and the production strategies that they engendered bothhave remained stable to the present day.

3.8. Agrarian dynamics, landscape transformation, and environmentalimpact

Figure 3 summarizes the main phases in the evolution of the agrarian systems ofthe Dao village of Ban Cuon and the Tày village of Ban Dieu, in terms of bothagrarian dynamics and related socioeconomic transformations. The land use mapsin Figure 4 reveal the environmental impact ofthose agrarian dy"namics. Between1983 and 1989, swidden fields were extended, beginning with the area around thecommune's paddyfields, dwellings, and roads. Hillside cultivation was muchmore important in the Tày villages than in the Dao village of Ban Cuon (Figure 5,north-west of the map). The Dao had already been cultivating the hillsides for

56

Farming / hunting / gathering

- Shlfting cullivation ln the mountains: 1

Dao and H'mong ethnic groups,long fallow periods

- Paddyfields managed collectivelyby the Tày

- Subslstence agnculture

Dao houseTày house

d!!J Kmh house

- Reduced fallow periods on slopinglands

- Natural resource management byindividual households

- Agricultural intensification onsloping lands near villagesterraces, hedges, etc.

- Lowlands no longer sufficient toensure food security (50 people/km')

- Land allocaled 10 farmers. Tàyreclaim ancestral land

- New village limits

- Rapid deterioration of forest cover

- Substantial erosion and acidificationof upland sOlls

- Rapid decrease in blodlversity

o Upland cropso Gardenso Paddy field

System limlls• Dense foresto Open forest

- Rapid upland deforestation

- Soil detenoration on sloping landsnear villages

- Floods destroy paddy fields nearthe river ln 1986 / 1996

- Sioping-land crops far from villagesto ensure food self-sufficiency

- Increasing population denslly: 25people/km'

- Cooperative system failing

- Lowland saturation- Collective lowland management,

private upland management

- Hillsides developed. in spite of bans

- Farming limlted to the lowlands- Collective natural resource

management

- Paddyfield intensification' improvedvaneties. manure. chemicalfertillzers

1 _ Abundant forest

- Lowland and valleys salurated

- High biodiversity ln the mounlalns

. - Slash-and-burn outlawed; privateforest use

; - Increasing population density fromnew immigrants from the lowlands

- Cooperative system

1960, _

Socioeconomic environment- Minimal contact between Tàyand

Dao ethnic groups

- Low population densily < 10people / km'

- Low level of market integratlon

Natural Resources

- Abundant forest

- Lowlands available for agriculture

- High biodiversity

~ Figure 3: Agricultura/ and socioeconomic transformations in Ngoc Phai

J.C. Castella. Tran Ouoe Hoa, O. Husson, Vu Hai Nam, Dang Dinh Ouang

1995

Legend

Paddy field_ Upland crop

Dense forest

Open forest

Shrub land

Grass land

n Other landNo information

Residential area

River 1stream

RaadCommune boundary

Source.' Aerial photographs for 1983 and 1989, SPOT satellite images for 1995 and 1998.

Figure 4: Land use maps of Ngoe Phai Commune for 1983, 1989, 1995, and 1998.

58

Rote of ethnicily in farm household differentialion

years, and cultural nonTIS opposed the clearing of any more land than wasnecessary. Thus, their practiccs on the hillsides did not substantially changeduring the Decree lOO period (1982-1988). Instead, in Ban Cuon there wassubstantial expansion of paddyiield area and a relative rcgelleration of forestland,charaeteristic of a slash-and-burn system with a long fallow period. The Tàyvillages, in contrast, had bcen eoncentrating 011 lowland tields for years, 50 forthem the availability and profitabillty of the hJllsides was new - they rushed toelear and elaim as ll1ueh land as possible.

8y the end of the 1980s, TLi.\' households were eultivating ail accessible arablesloping lands in their villages (Figure 6). Natural forest eover (forcst and shrubland from Figure 6-A) reached its lowest level in the history of Ngoc Phai. Theupland rush naturally ended with the exhaustion of uplancl forcst areas, and wassaon followcd by the Tù\' movement ta reelail-,) ancestral paddvLlllds. v\'hicheffectively reversed the Tàv/Doo agrarian dYll(lInic~ of' the \nos. As thc Tà\ inthe south of the commune focused on thcir ne\\ Iy-acquirecl paddy fielcls, the forestregenerated. Meanwhile .. deprivcd oi' their paddy land fields, the Dao rcsorred taincrcascd sloping-Iand cultivation (Figure 6-8) with shortening fallow perioclsand more land being croppecl continuously,

Recently, farming lanclseapes in TLiy villages have reaehed a sustainablc balance,while the environll1ental degradatioll around the Dao village IldS bcen aggravatcdby a large-scale return to sJash-and-burn systems. Sueh sy_tcms, driven by thelack of aeeess ta paddy lands, arc steadily destroying the resource base on whichthey draw, and local farmers are looking for alternatives (Castella ct al., 2002),

"",0 2

km

Legend

Paddy field increase

_ Upland crop Jncrease

_ Forest recovering & reforestalion

Shrub increase

Grass land increase

._. Forest degradation

Unchanged

_ SeUlement expansion

River 1 stream

Raad

Commune boundary

Source. 1983 and 1989 land use maps in Figure 4.

Figure 5: Land use changes in Ngoc Phai Commune between 1983 and 1989.

59

J. C Castella, Tran Quoc Hoa, O. Husson, Vu Hai Nam Dang Dinh Quang

400

600

Area (ha)

8000~

~

r rio

500

1500

1000

Area (ha)

2500

2000

1983 1989 1995 1998 1983 1989 1995 1998

Denseforest

o Open D Shn~forest lan~

o Paddyfield

o Uplandcrop

o Grassland

Figure 6: Land use changes. A: Forest and natura! vegetation, B: Farmland

3.9. The evolution of animal hûsbandry practices

Animal husbandry in the commune expanded both as a means of storing capitalfor farmers and as a response to the recent development of the meat market for thedelta zones (Figure 7) The steep increase in numbers of large ruminants (buffaloand cattle) has created a threat to sloping lands and forests. The transition fromcollective herd management to management by individual households has led tofree-grazing practices in both the lowlands and hillsides, resulting in an increascdnumber of disputes between farmers (Eguicnta et al., 2002). Livestock ownersreturned to this traditiona! practice becausc individual households did not haveenough labor force to tend a few head of livestock ail year round.

Farmers have developed a number ofsoJutions to the problems caused by roaminganlmals:

- Bui Iding fences (of wood or bamboo) for the short tcrm and then planting hcdgesand digging trenches around cultivated plots;

- Temporarily (during the crop cycle) establishing shelters close to agriculturalplots to better survey crops;

- Growing crops in areas inaccessible to animais.

[n Ban CUOIl, the existing uplar.d erop systems deterl11ined the methods ofbuffalomanagement. The animais were lllonitored during the day and stabled in theevenings throughout the whoJe up[and rice-growing season. Given that a specifieplot could be planted to upland rice for no more than 4 successive years, and thattheir areas were often large, Dao farmcrs were not \Villing to invest in hedges andtrenches. In contrast, terraced paddyficlds can be cropped permanently. Thus theusers often built barriers around their terraced paddyfields.

60

Role of ethnicity in farm housE·hold differentiation

Figure 7: Plg and buffalo numbers in Ngoc PhaiCommune

Source: Cho Don District's Office of Agricultuream: Rural Development

2000989694

Year

9290o

No. of head

1.200 .~._~-

800 ! ~ ....!().._,~~~ ..... -.--_',__'----,

o plgs 1

o buffaloes '

In the Tày villages, swiddencrops were systematically pro­tected with hedges and trenches.AnimaIs were rarely monitored,even during the growing season.Sioping land cultivation thusrequired an increased invest­ment (fence building or plantinghedges) if crops were to beprotected from roaming ani­mais. Tày farmers often chosesimply not to cultivate thoseareas where buffaloes werepresent in large numbers.

The free-grazi ng practices of Tày farmers Ilave not soothed relations betweenethnie groups in the region. Tày previously had abruptly removed the Dao fromtheir paddyfields by the ancestral land reciamation movement, and now Tàylivestock are causing a major problem for the Dao's upland harvests. Researchersare looking for solutions that will undoubtedly require new spatial arrangements toreconcile crop with livestock management on the hillsides (Castella et al., 2002a).

4. Farm diversity and access to the means of production

The previous section demonstrates that historically, ethnicity has been a majordeterminant of paddyland access and household differentiation. But in present­day Bac Kan, the existing diversity in land use and access cannat be attributedexclusively to ethnicity. At present, the comparative advantage of paddy land ricerelative ta lIpland rice has made it the priority for farmers of ail ethnicities. Wherethe possibility has existed, Dao farmers have established permanent settlementsand purchased paddy fields. Conversely, many Tày farmers have been forced toturn to shifting cultivation as a means ofsurvival when their paddy land access wasrestricted (e.g. sons inheriting from their parents paddy area not suffïcient to feedtheir family). Production systems today are no longer divided along ethnic lines.

4.1. Means of access to paddyfields

The disputes callsed by the unequal distribution of the paddylands dissipatedtoward the end of the 1990s, as land-deprived fanners took stock of the final ity ofthe 1993 land law. That they did not seek to overturn the law indicates that accessto paddyfields had been defined relatively clearly and accepted by this time.There remained several means of access to paddy fields.

61

J.C. Castella, Tran Ouoe Hoa, 0. Husson, Vu Hai Nam, Dang Dinh Ouang

Pure hase of paddyland

ln Ngoc Phai Commune, the first land sales took place in 1992. The Tày who wererelocated to the interior of the commune in the 1962 resettlement were the first tosell off their land, either in part or in its entirety. The first paddyfields sold were oflow quality: either terraced or in places with pOOl' soil, insufficient water, orinsufficient solar radiation (e.g., deep narrow valleys). In Ban Cuon, the majorityof the Tày chose to sell their l'icefields to Dao farmers for the following reasons:

- The desire to unite separated families: while parents often continued to live inneighboring Ban Dieu village, their children were managing lands in Ban Cuon.

- Crop dispersion: farmers owned paddyfields in Ban Cuon, and maize fields atBan Dieu. By consolidating ail their holdings in a single area, their productioncould become more efficient (Jess time wasted travelling between sites).

- Yields at Ban Cuon were pOOl' in comparison with the l'est of the commune(resulting from the recent development of the paddyfields).

The Tày founders, after rcclaiming their ancestral land, were not interested inreselling it. But between paddyfields and sloping-land swidden crops, they soonfound that they had more land than they could make use of with their labor force.Theil' most practical option was a single annual rice crop on paddy land. In termsofefficiency and production pel' se, it would have made sense to sell off the uplandand focus on paddy intensification. Nonetheless, they did not take that optionbecause by continuing to produce swidden crops, farmers could maintain propertyrights on the cleared land. Nor did thcy want to divest themselves of anypaddyland. Paddy lands were the most important component of the productionsystem, and were held to kcep land security for future descendants. Therefore,land sales by descendants offounding families have begun to take place only veryrecently. The sellers tend to be families with insufficient labor to maintain theirpaddyfields, and a small number of descendants. The profits from sales ofpaddyland are generally reinvested in new equipment (huskers, hand tractors),livestock (cattle), and cement houscs.

By the year 2000, it had become very difficult to find land to purchase in the Tàyvillages, though land transactions continued to take place in Ban Cuon. Between1992 and 2000, the priee of paddyfields rose by a fàctor offour (Figure 8).

Inheritanee of paddy fields

A young married couple could gain access to a paddyfield by inheriting a portionof a parent's land. In the Tày communities, land traditionally was passed fromfather to son, keeping the land in the line of male descent and preventing anoutsider from gaining access to land through marriage. The Dao only recentlyadopted this principle, having never before been landowners. With the growingshortage of paddytields and the Dao's difficulties in obtaining them, they havebecome very attached to patriarchal land inheritance.

62

Role of ethnicity in farm household differentiation

Ban Cuon Ban Dieu

D 1992 1995 D 1999

Borrowing and lending ofpaddy fields

Lending most often takes placewithin a family, between bro­thers or between father and son.However, a farmer occasionallywill lend land to a farmer inanother family, if the lenderowned more than he was able tocultivate. The lender generallydoes not receive any compen­sation for the loan.

Priee in million VND

8

6

4

2

o -

Source: Fietd surveys.

Figure 8: Paddyfietd price changes in two villages inNgoc Phai.

NB.: Average price of a 1000 m2 paddyfietd plotadapted for doubte-cropping in 1999 VND.

Tenant farming

Rarely practiced in the com­mune, but recent examples illus­trate the growing interactionbetween ethnie communities.At Ban Cuon, some Dao farmers benefited in 1999 from the temporary use of aPllddyfield belonging to Tày farmers. The Dao t'armers were allowed to work theland during the spring and reap the harvest. In exchange, they prepared the soil(plowing, harrowing) and provided buffalo manure for the second crop, which wascultivated by the Tày landowners.

Renting

We did not observe land renting over the course of our survey, but it existed in thecommune between 1993 and 1995. Tenants had to give landowners a sum ofmoney or portion of the harvest in retum for the use of the land. According to theformer tenants interviewed, this practice did not last long, as landowners preferredto sell their land.

Building new paddyfields

The only households who could engage in the task, particularly in Ban Clion, werethose with both many laborers (or enough capital to hire help), and land suitablefor development (close enough to a water source for irrigation, few stumps andstones, etc.).

4.2. Means of access to sloping lands

ln the early 1990s, forestlands were allocated to households in such a way thatmost farmers received the rights for whatever lands they were cultivating at thetime (Castella et al., 2002). In the swidden expansion of the 1980s, the foundingTày families, Jike the Dao, had cleared the sloping lands nearest their settlements.

63

J.C. Castella, Tran Ouoe Hoa, O. Husson, Vu Hai Nam, Dang Dinh Ouang

This forced fàmilies arnvIng laler in the collectivization to go farther away(beyond village limits) to find fertile land, Many such households did not claimthese lands when forestlands werc being allocated in the early 1990s. Sorne whohad the option chose not to act on it, as the land was too far from their residences.There was little benefit in becoming the owner of land that was either unusable orimpossible to monitor. Still others were dissuaded by the fear that a tax would beimposed on the new allotments.

Unsanctioned use of common land

ln the early 1990', local authorities categorized uplands and specified their uses(Mellac, 1997; Castella et al., 2002). However, fàrmers mostly ignored thesespecified categories. Many farmers continued to cultivate commons supposedlyreserved for pastures, protected forests, or zones at the bases ofhillsides. Farmersconcentrated first on common land because (i) they each had signed agreements toprotect their allocated upland, whereas the protection of common land was lessclearly defined; and (ii) by clearing common land, they hoped that it might oneday become theirs. Many farmers realized this hope, as a second allocation in1997 indeed gave to fàrmers the common lands that they were cultivating.

Although common land played a relatively small role in Tày production, it wasessential for the struggling Dao households, who relied on upland cultivation.Avoiding cultivation on their allocated lands wherever possible because they hadsigned protection agreements, Dao farmers turned to:

(i) Communal pasture lands. In Bali Cuon, hillside crops were cultivatedintensively in zones initially set aside for animal pasture. In 1999,approximately 60% of houscholds were cultivating upland rice on landofficially designated as pasturelands. Local authorities temporarily authorizedswidden crops in order to expand the pastureland area. Access to these landsseemed to follow no particular mIe: the land cou Id be cultivated even withoutauthorization from the village headman.

(ii) Land in other villages or communes. In 1999, ten Dao families from BanCuon fàrmed Tày-owned land outside of Ngoc Phai. Authorities of theneighboring communes eventually put an end to these slash-and-burnpractices after several disputes between the Tày owners and Dao occupants.

Borrowing and purchase

A few isolated land transactions did take place between Ban Cuon farmers andneighboring villages. One Dao fàrmer "borrowed" a plot of land that had beenused in the previous year by another Dao fàrmer from a neighboring village whohad decided to abandon il. We came across only one case of an upland purchaseover the course of our research in Ngoc Phai. A Dao farmer purchased a swiddenplot of sorne 3,000 m2 for 800,000 VND from a Tày farmer of a neighboringcommune who had been using the land. This kind of transaction is more

64

Role of ethnicity in farm household differentiation

developed in other communes of Bac Kan with more severe land scarcity (Altheret al., 2002). Renting and tenant farming do not exist in the uplands.

Allocated lands covered by forest-protection contracts

Several farming approaches were observed on allocated lands:

(i) A fanner burns a single plot, out ofsight from forest wardens (not visible fromthe road).

(ii) Under the authority ofa forest and fruit tree plantation program (e.g., nationalreforestation program 327, or international development projects), some plotsare cleared completely, then planted with seedlings of the desired tree species.This allows several years of planting annual crops between the young trees.Annual cropping ends when the trees grow big enough to block thc sunlight.

(iii) A farmer burns wherever he determines that the soil is weil adapted to his cropchoice, without regard to the forest wardens. In the interest of supportingagriculture, forest wardens keep fines relatively low (Zingerli et al., 2002).

(iv) Since 1999, several poor Ban Clion tànners requested special permission toengage in slash-and-burn practices on their allocated land, as a last resort tofeed their households. The commune's forest wardens either granted ordenied the tàrmer this right, based on the quantity and quality of land that hepossessed. If the request was granted, the forest wardens then regulated thepractice, specifying the area and location of the burn.

4.3. Househo/d typ%gy and Iivelihood strategies

Based on the access mechanisms described above, several different householdtypes have developed in Ngoc Phai, each with its own particular productionstrategy. Figure 9 shows the differentiation trajectory that gave rise to eachhousehold type. Table 3 identifies the characteristics of the resulting householdtypes and Figure 10 graphically demonstrates their associated productionsystems. The classification of households permitted us to better understandtànners' current situations and the reasons behind their current livelihoodstrategies. In this section, we will describe the various production alternativesemployed by the various household types

The products of crop and livestock production

At the level of the commune, the main commercial products from crop productionwere rice, soybeans, and maize. The owners of the largest paddyfields(households in Types 1 and II) often sold upland rice surpluses. Natural1y, BanClion tànners without paddyfields (Type VI) did not sell any upland rice, becausethey used any upland production for family consumption. Ail householdspossessed at least some pigs. Crops with harvests spread across the year, likecassava, were generally grown to feed pigs (ail Types) and the sale of animaIs was

65

J.C. Castella, Tran Quoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Quang

often a significant income source. Depending on the particular husbandry strategy(breeder-fattener or fattener), revenue from pig husbandry cou Id be eitherimmediate or spread out across the year.

In Ban Cuon, sorne Dao farmers (Type IV) were able to raise funds to repurchasepaddyfields from founding Tày families by selling their buffaloes. Ownership ofa substantial number of livestock (5 head or more) at the time when the firstpaddyfields were being sold by the Tày was the major determinant ofdifferentiation in Ban Cuon.

The raising of large ruminants (buffaloes and cattle) was an important means ofcapitalization and long-term saving, whereas pig farming was a medium-termsource of income for timely investment needs (expenses related to festivals,chemical input purchases, etc.).

paddyfield collective period upland resource land marketarea/laborer-----.......f---------------.. .. exploitation .......f-----.....

1960 1970 1980 1990 2000

l'

A1: Tày founding familiesA2: "Immigrant families" (Dao, Tày, Kinh,Nùng)

81: Tày cooperatives

82: Dao cooperatives

C1: Tày, Kinh farms, small labor force, smallswidden areas, food self-sufficiency

C2: Tày, Kinh farms, large labor force, largeswidden plots, surpluses

C3: Dao farms, smal"abor force, averageswidden areas, few buffaloes, food self­sufficiencyC4: Dao farms, large labor force, very largeswidden areas, many buffaloes, surpluses

1: Founding Tày families with reclaimed land, ricesurpluses, many buffaloes

Il: Young Tày couples (descendants of foundingfamily), recently settled on inherited paddyfields, riceself-sufficiency, maize sales, few buffaloesIII: "Immigrant" families, whose numbers grew in the80s, many buffaloes, ricefields purchased in 1993,rice self-sufficiency.

IV: Tày founding families no longer living in theirhome villages, repurchased paddyfields, riceshortages, substantial non-farm activities.

V: "Immigrant" families who recently repurchasedpaddyfields, numbers growing during the 90s, riceshortages, large swidden areas.

VI: "Immigrant" families without paddyfields, riceshortages, few buffaloes, large swidden areas (maizeand upland rice), substantial non-farm practices.

Figure 9: Househo/d different/at/on patterns in Ngoc Phai Commune

66

Table 3: Main characteristics of household types identified in Ngoc Phai

Farm type 1 1/ II/ IV V VI

ArrivaI date founding familiesdescendants of

founding families at the beginning of and during the cooperative periodfounding families

Fami1y composition6 to 7 members 4 to 5 members 6 to 7 members 6 to 7 members 4 to 7 members 5 members

31aborers 21aborers 3 laborers 3 laborers 2 to 3 laborers 3 laborers

Means of access toreclamation inheritance

reclamation andpurchase in 1993 no paddyfields recent purchase

paddytields purchase

Paddy area 1laborer2000-2400 1500-2000 500-700 700-1200 0 500-700

(m2)

Allocated forest area 3-5 ha 1.5-3 ha 1.5 ha 10-\2 haDao (Va): 5-10 ha

1.5-3 haTày (Vb): 1 ha

Rice self-sufficiencyself-sufficientl

self-sufficient shortages self-sufficient shortages shortagessurpluses

Sale of swiddenupland rice maize, up1and rice

productionno malze no no

Annual buffalo sale yes no no yes no no(no. ofheads owned) (6) (2) (2) (5 - 7) (0 - 3) (2)

Type of pig raising intensive intensive intensive semi-intensive light(Va,I-6) intensive(no,ofheads) (3 - 6) (2) (4) (8-12) intensive (Vb, 4-6) (3 )

Size ofperenniallarge medium small large small medium

plantations

1mportance of non-unimportant important important important

unimportant (Va)unimportant

farrning activities important (Vb)

Net revenue 1 laborer 14,820 4,270 3,020 4,050 2,240 2,590

year (x 1000 VND)(% agric. revenue)

( 100) (80) (60) (75) (Va=85, Vb=45) (90)

J. C. Cas/ella. Tran Ouoc Hoa. O. Husson, Vu Hai Nam, Dang Dinh Ouang

Capital

ype Va Paddy rice self­sufficiency lhreshold

Type 1

Paddy field area

Figure 10: Op/ions for income genera/ion based on paddyfie/d area and capital accumulation.

Non-farm income

ln Ngoc Phoi, the majority of f,lI'mers engaged in non-farm aetivilies. The limegiven to these activities depended on the houschold type. The households thatmade substantial investments (in time and labor) in non-farm activities were:

(i) households with little or no paddyfield area (Types V and Vl),

(ii) households with a high laborer/cultivated area ratio (Types III, IV, V and VI),and

(iii) households with access to forest resources for commercial exploitation.

Principal non-farm activitics included:

- Fores! resollrce lise. The primai)' non-agricultural activity was the gathering offorest products (Types Il, nI and IV), including sales of firewood, balllboo,bamboo shoots, rattan, and SOIllC kinds of wild animais (snakes, turtles, androdcnts).

- Commerce, Village shops fell into two categories: (i) households withoutpaddyfields, able to dedicalc a large part of rheir rime to commerce (Type Vb);and (ii) households with surplus labor force and sufticient capital to open a shop

68

Raie af ethnicity in farm hausehald differentiatian

while concurrently pursuing their farrning activities (Type III). It is often theelderly women who take care of the shop.

- O/her activities. Non-agricultural revenue also was generated from serviceprovision, such as house construction, brick-making, or motorcycle taxi.

5. Conclusions: strategies based on access to productionmeans, not ethnicity

Historically, the two major ethnic groups populating Bac Kan Province inhabitedseparate tiers of the ecosystem. The sedentary Tày cultivated paddy rice in thevalley-bottoms, while the nomadic Dao harvested rainfed rice on swidden fieldsin the uplands. The two groups rarely mixed. Historically, the Tày received themajority access to lowland fields by privilege oftheir ancestral rights, and the Daowere left to find their livelihoods in the forest.

Since the beginning of Vietnamese independence, several State policies havealtered the institutional environ ment in which mountain people live, and havecaused various levels of social differentiation between and within the ethnicgroups. At present, both groups follow similar livelihood strategies, basingproduction decisions on the relative profitability of land and labor dedicated tovarious activities. These strategies hinge on one crucial factor: access to paddyfields (Castella and Erout, 2002). Where land purchases are possible, we seetraditionally-migratory Dao farrners willingly taking up sedentary living, to takeadvantage of the stability of irrigated rice production. Conversely, wheretraditionally-sedentary Tày farmers do not have access to lowland fields, Tàyfarrners have tumed to the shifting cultivation systems traditionally practiced bythe Dao. It is now becoming difficult to draw clear lines between the agriculturalpractices and lifestyles of these two major ethnic groups in Bac Kan Province.Instead of using the traditional criterion of ethnicity, researchers should usehouseholds' endowments of land, labor, and natural resources as the key factorsfor analyzing current circumstances and future rural development actions.

Acknowledgements

The authors are very grateful to Paul Novosad for translating this chapter from aprevious French version and for his contribution in further editing the Englishversion of this chapter.

69

J.C. Castella, Tran Quoc Hoa, O. Husson, Vu Hai Nam, Dang Dinh Quang

References

Alther c., Castella J.c., Novosad P., Rousseau E. and Tran Trong Hieu (2002) Impact ofaccessibility on the range of livelihood options available to farm households inmountainous areas ofnorthern Viet Nam. In: (J.c. Castella and Dang Dinh Quang eds.)Doi Moi in the Mountains. Land Use Changes and Fm'mers' Livelihood Strategies inBac Kan Province, Viet Nam. The Agricultural Publishing House, Ha Noi, Viet Nam.121 - 146.

Bal P., Mellac M. and Duong Duc Vinh (1997) Evolutions récentes des systèmes deproduction dans une zone de montagne du Nord-Vietnam, district de Cho Don, provincede Bac Kan. Cahiers Agricultures, 6 and Agriculture et Développement, 15: 183-189.

Bal P., Castella J.c., Le Quoc Doanh, Husson O., Tran Dinh Long, Dang Dinh Quang, HaDinh Tuan and Duong Duc Vinh (2000) Diagnostic systémique, rechercheagronomique et appui au développement: exemple d'une intervention concertée dans laprovince de Bac Kan. In: (VASI-GRET eds.) Appui à l'Organisation de la ProductionAgricole dans le Nord du Vietnam. The Agricultural Publishing House, Ha Noi, VietNam. 57 - 92.

Castella lC., Boissau S., Nguyen Hai Thanh and Novosad P. (2002) Impact of forestlandallocation on agriculture and natural resources management in Bac Kan Province, VietNam. In: (lC. Castella and Dang Dinh Quang eds.) Doi Moi in the Mountains. LandUse Changes and Farmers' Livelihood Strategies in Bac Kan Province. Viet Nam. TheAgricultural Publishing House, Ha Noi, Viet Nam. 197 - 220.

Castella J.c. and Erout A. (2002) Montane paddy rice: The cornerstone of agriculturalproduction systems in Bac Kan Province, Viet Nam. In: (J.c. Castella and Dang DinhQuang eds.) Doi Moi in the Mountains. Land Use Changes and Farmers' LivelihoodStrategies in Bac Kan Province, Viet Nam. The Agricultural Publishing House, Ha .1I/oi,Viet Nam. 175 - 195.

Castella lC., Tran Trong Hieu and Eguienta Y. (2002a) Crop-livestock interactions innorthern Viet Nam uplands. II. Use of spatial graphie models as a common languagebetween researchers and local stakeholdcrs. SAM Paper Series 12, VietnamAgricultural Science Institute, Ha Noi, Viet Nam. 17 p.

Castella lC., Tronche N.R. and Vu Nguyen (2002b) Landscape changes in Cho DonDistrict since the doi moi era (1 990-2JOO) and their implications for sustainable naturalresource management in Viet Nam's mountainous provinces. In: (J.c. Castella and DangDinh Quang eds.) Doi Moi in the Mountains. Land U5e Changes and Farmers'Livelihood Strategies in Bac Kan Province, Viet Nam. The Agricultural PublishingHouse, Ha Noi, Viet Nam. 149 - 173.

Dang Nghiem Van (1991) La culture sur brûlis et le nomadisme. Etudes Vietnamiennes,1(99): 16-28.

De Koninck R. (1999) Deforestation in Viet Nam. International Development ResearchCentre, Ottawa, Canada. 100 p.

70

Raie af ethnicity in farm hausehald differentiation

Eguienta Y.K., Martin c., Lecomte P., Husson O. and Castella lC. (2002) Crop-livestockinteractions in northern Viet Nam: Issues, diversity of farmers' responses andalternatives for sustainable integration of animaIs in upland agricultural systems. In:(lC. Castella and Dang Dinh Quang eds.) Doi Moi in the Mountains. Land UseChanges and Farmers' Livelihood Strategies in Bac Kan Province. Viet Nam. TheAgricultural Publishing House, Ha Noi, Viet Nam. 221 - 247.

Husson O., Castella J.c., Ha Dinh Tuan and Naudin K. (2001) Agronomie diagnosis andidentification of factors limiting upland rice yield in mountainous areas of northcrnVietnam. SAM Paper Series 2, Vietnam Agricu1tural Science Institute, Ha Noi, VietNam. 16 p.

Mellac M. (1997) L'état et la forêt au Nord Viet Nam. Cahiers d'Outre Mer, 50: 27-42.

Morrison E. and Dubois O. (1998) Sustainable Livelihoods in Upland Vietnam: LandAllocation and Beyond. Forestry and Land Use Series 14, International lnstitute forEnvironment and Development (lIED), London, England. 64 p.

Sadoulet D., Castella J.c., Vu Hai Nam and Dang Dinh Quang (2002) A short history ofland use changes and farming system differentiation in Xuat Hoa Commune, Bac KanProvince, Viet Nam. In: (lC. Castella and Dang Dinh Quang eds.) Doi Moi in theMountains. Land Use Changes and Farmers' Livelihood Strategies in Bac KanProvince. f'let Nam. The Agricultural Publishing House, Ha Noi, Viet Nam. 21 - 46.

Tran Van Ha (1993) Histoire des Transformations de l'Agriculture au Viet Nam 1945­1991. Cas de la Commune de Dong Vien. Programme Fleuve Rouge, GRET-VASI, HaNoi, f'let Nam. 36 pp.

Zingerli c., Castella .I.e., Pham Hung Manh and Pham Van Cu (2002) Contesting policies:Rural development versus biodiversity conservation in the Ba Be National Park area,Viet Nam. In: (J.c. Castella and Dang Dinh Quang eds.) Doi Moi in the Mountains.Land Use Changes and Farmers' Livelihood Strategies in Bac Kan Province. Viet Nam.The Agricu1tural Publishing House, Ha Noi, Viet Nam. 249 - 275.

71

DoiMoiin the Mountains

Land use changesand Carmers' livelihood strategiesin Bac Kan Province, Viet Nam

Edited byJean-Christophe Castellaand Dang Dinh Quang

The Agricultural Publishing HouseHa Noi, Viet Nam2002

Vietnam Agricultural Science InstituteVASI, Thanh Tri, Ha Noi, Viet Nam

Institut de Recherche pour le DéveloppementIRD, 213 rue Lafayette, 75480 Paris Cedex 10, France

International Rice Research InstituteIRRl, DAPO Box 7777, Metro Manila, Philippines

© VASI - IRD - IRRI, 2002

Jean-Christophe Castella and Dang Dinh Quang (eds.) 2002Doi Moi in the Mountains:

Land use changes and farmers' livelihood strategies in Bac Kan Province, Viet Nam1. Viet Nam. 2. Rural development. 3. Mountain agriculture

Published in Ha Noi, Viet Namby

The Agricultural Publishing House

Publishing Permit No. 11861dated 21106/2001